Question 50
A team maintaining an MCP integration server must meet 2 separate acceptance criteria: (1) offer users a reusable parameterized instruction template from an MCP server; (2) make a repeatable multi-step code-review methodology available to an agent. Which two implementation actions directly satisfy the stated requirements? Choose TWO.
- Create a Skill containing the procedure and supporting instructions.
- Register a prompt template with its required arguments.
- Expose the material as an MCP resource with a stable identifier.
- Implement a client-side handler and return its result with the tool-use identifier.
- Put an approval gate in the host-side write handler.
Explanation
Correct Answers
A:The test to make a repeatable multi-step code-review methodology available to an agent is met because a Skill teaches how to perform a task, whereas a tool is the executable capability that performs an external operation; the remaining selection covers the separate need to offer users a reusable parameterized instruction template from an MCP server.
B:The test to offer users a reusable parameterized instruction template from an MCP server is met because MCP prompts provide structured prompt content, but by themselves do not perform external business actions; the remaining selection covers the separate need to make a repeatable multi-step code-review methodology available to an agent.
Incorrect Answers
C:The action is useful for the separate capability to make reference material available for clients to read as context because resources represent consumable data, contrasting with tools that invoke an operation or prompts that package reusable instructions; however, the required test outcomes are to make a repeatable multi-step code-review methodology available to an agent and to offer users a reusable parameterized instruction template from an MCP server.
D:The action is useful for the separate capability to connect a Claude tool request to an internal service whose logic the application owns because a declared client tool does not execute itself; application code runs it and feeds results back into the conversation; however, the required test outcomes are to make a repeatable multi-step code-review methodology available to an agent and to offer users a reusable parameterized instruction template from an MCP server.
E:The action is useful for the separate capability to ensure an agent obtains explicit authorization before updating a customer record because an irreversible or sensitive write must be permitted by trusted application logic, not merely because the model selected a tool; however, the required test outcomes are to make a repeatable multi-step code-review methodology available to an agent and to offer users a reusable parameterized instruction template from an MCP server.
Question 51
A team maintaining a Claude API workload benchmark has two different engineering requirements: (1) stop interpreting non-2xx API responses as normal assistant messages, and (2) reject incomplete or malformed response data before business logic consumes it. Which TWO independent changes directly meet them? Choose TWO.
- Branch on HTTP status and decode the provider's error structure.
- Benchmark an Opus-family model on the hardest cases.
- Use a WebSocket protocol only when the target endpoint actually offers one.
- Keep HTTPS certificate and hostname verification enabled.
- Parse structured payloads with explicit failure handling.
Explanation
Correct Answers
A:To stop interpreting non-2xx API responses as normal assistant messages as part of a design that must also reject incomplete or malformed response data before business logic consumes it, transport errors are not regular model completions, so application control flow must treat them separately; this is why the action is one of the two required selections.
E:To reject incomplete or malformed response data before business logic consumes it as part of a design that must also stop interpreting non-2xx API responses as normal assistant messages, JSON parsing validates syntax, while separate schema and semantic checks verify types and permitted values; this is why the action is one of the two required selections.
Incorrect Answers
B:This alternative serves the separate goal to solve demanding multistep reasoning tasks where quality matters more than lowest unit price because higher-capability tiers may improve difficult-task performance, but their cost and latency still need objective justification; the required outcomes are instead to stop interpreting non-2xx API responses as normal assistant messages and to reject incomplete or malformed response data before business logic consumes it.
C:This alternative serves the separate goal to build bidirectional persistent event communication where the chosen platform supports it because webSockets can carry two-way events over a long-lived connection; they are not a drop-in replacement for every HTTP or SSE endpoint; the required outcomes are instead to stop interpreting non-2xx API responses as normal assistant messages and to reject incomplete or malformed response data before business logic consumes it.
D:This alternative serves the separate goal to protect model API traffic against an impersonated network endpoint because verified TLS authenticates the service and protects data in transit; disabling validation defeats a key transport security property; the required outcomes are instead to stop interpreting non-2xx API responses as normal assistant messages and to reject incomplete or malformed response data before business logic consumes it.
Question 52
An operational review of a specialist-agent orchestration service reports distinct failures: it cannot minimize infrastructure ownership for a supported hosted agent execution surface, and it cannot restrict a development agent to a defined set of file and command capabilities. Which two repairs are required? Choose TWO.
- Apply limits for steps, elapsed time, and token or tool expenditure.
- Configure tool permissions rather than depending only on a prompt.
- Configure the appropriate Anthropic-executed server tool.
- Evaluate a managed agent runtime against the deployment requirements.
- Consume the agent runtime's streamed event sequence.
Explanation
Correct Answers
B:To restrict a development agent to a defined set of file and command capabilities as part of a design that must also minimize infrastructure ownership for a supported hosted agent execution surface, runtime permissions constrain what actions can run even if the model requests more access; prompt instructions alone are not a security boundary; this is why the action is one of the two required selections.
D:To minimize infrastructure ownership for a supported hosted agent execution surface as part of a design that must also restrict a development agent to a defined set of file and command capabilities, a managed runtime can take on execution infrastructure, but access controls, product availability, and data-handling needs still require evaluation; this is why the action is one of the two required selections.
Incorrect Answers
A:This alternative serves the separate goal to impose predictable maximum cost and wall-clock exposure on autonomous execution because runtime budgets are external constraints on agent activity and must be enforced by the host, not inferred from a cooperative natural-language instruction; the required outcomes are instead to minimize infrastructure ownership for a supported hosted agent execution surface and to restrict a development agent to a defined set of file and command capabilities.
C:This alternative serves the separate goal to use a Claude platform-provided capability without operating a local handler because server tools execute on the provider side and have distinct authorization and runtime semantics from client tools the application dispatches; the required outcomes are instead to minimize infrastructure ownership for a supported hosted agent execution surface and to restrict a development agent to a defined set of file and command capabilities.
E:This alternative serves the separate goal to deliver intermediate agent events to the client as they occur because event-driven consumption lets the application display progress and inspect tool actions without waiting for the entire multi-step run to finish; the required outcomes are instead to minimize infrastructure ownership for a supported hosted agent execution surface and to restrict a development agent to a defined set of file and command capabilities.
Question 53
While planning a function-calling service, the product owner specifies that it must require or restrict tool use for a specific turn when the API supports that control. Which solution fulfills the requirement?
- Configure the built-in capability instead of recreating it as a local handler.
- Configure the supported tool_choice behavior for the request.
- Implement a client-side handler and return its result with the tool-use identifier.
- Write a custom tool whose handler runs in the application.
Explanation
Correct Answer
B:To require or restrict tool use for a specific turn when the API supports that control in this function-calling service, tool-selection controls influence whether Claude may call tools, but cannot grant permissions to the underlying systems; the proposed implementation directly supplies that capability.
Incorrect Answers
A:The proposed alternative helps to obtain a capability already provided as a supported Anthropic server tool because provider tools simplify execution ownership, though availability, cost, and data handling still require assessment; that mechanism does not meet the different requirement to require or restrict tool use for a specific turn when the API supports that control.
C:The proposed alternative helps to connect a Claude tool request to an internal service whose logic the application owns because a declared client tool does not execute itself; application code runs it and feeds results back into the conversation; that mechanism does not meet the different requirement to require or restrict tool use for a specific turn when the API supports that control.
D:The proposed alternative helps to invoke a proprietary internal pricing database unavailable through a built-in feature because custom tools expose application-specific actions while retaining control over authentication, validation, and network access; that mechanism does not meet the different requirement to require or restrict tool use for a specific turn when the API supports that control.
Question 54
Two approval criteria are still open for a Claude inference cost review: the team must serve high-volume straightforward classification at a tight latency and cost target and preserve critical service during a temporary primary-model problem. Which two implementation actions satisfy them? Choose TWO.
- Branch on HTTP status and decode the provider's error structure.
- Iterate the SDK or HTTP event stream until its terminal event.
- Design an approved fallback model path with compatibility checks.
- Benchmark a suitable Haiku-family model for the simple task.
- Benchmark an Opus-family model on the hardest cases.
Explanation
Correct Answers
C:To preserve critical service during a temporary primary-model problem as part of a design that must also serve high-volume straightforward classification at a tight latency and cost target, fallbacks improve resilience only when downstream schemas, cost limits, and policy controls work with the alternate model; this is why the action is one of the two required selections.
D:To serve high-volume straightforward classification at a tight latency and cost target as part of a design that must also preserve critical service during a temporary primary-model problem, lightweight model tiers can favor speed and unit economics, but accuracy must still meet the specific evaluation threshold; this is why the action is one of the two required selections.
Incorrect Answers
A:This alternative serves the separate goal to stop interpreting non-2xx API responses as normal assistant messages because transport errors are not regular model completions, so application control flow must treat them separately; the required outcomes are instead to serve high-volume straightforward classification at a tight latency and cost target and to preserve critical service during a temporary primary-model problem.
B:This alternative serves the separate goal to process partial model output as it is delivered over a streaming connection because streaming clients must handle both content deltas and final status rather than assuming the first chunk is the complete response; the required outcomes are instead to serve high-volume straightforward classification at a tight latency and cost target and to preserve critical service during a temporary primary-model problem.
E:This alternative serves the separate goal to solve demanding multistep reasoning tasks where quality matters more than lowest unit price because higher-capability tiers may improve difficult-task performance, but their cost and latency still need objective justification; the required outcomes are instead to serve high-volume straightforward classification at a tight latency and cost target and to preserve critical service during a temporary primary-model problem.
Question 55
A team maintaining a machine-readable extraction API must meet 2 separate acceptance criteria: (1) safely render generated text in a browser without executing embedded markup; (2) stop verbose tool payloads from consuming most of a long-running agent's context. Which two implementation actions directly satisfy the stated requirements? Choose TWO.
- Put stable application constraints in the system instruction field.
- Retain essential fields and discard irrelevant tool output from active context.
- Supply representative labeled examples including edge conditions.
- Escape or sanitize model text in the output presentation layer.
- Revise one prompt component at a time against a regression evaluation set.
Explanation
Correct Answers
B:The test to stop verbose tool payloads from consuming most of a long-running agent's context is met because summarizing or filtering observations keeps the reasoning window focused without deleting required facts from durable application state; the remaining selection covers the separate need to safely render generated text in a browser without executing embedded markup.
D:The test to safely render generated text in a browser without executing embedded markup is met because generated content is untrusted for HTML rendering; injection into the DOM must not become executable client-side code; the remaining selection covers the separate need to stop verbose tool payloads from consuming most of a long-running agent's context.
Incorrect Answers
A:The action is useful for the separate capability to apply the same compliance restrictions across many different end-user requests because high-priority application instructions should not be delivered as ordinary user content where they compete with user questions and document text; however, the required test outcomes are to stop verbose tool payloads from consuming most of a long-running agent's context and to safely render generated text in a browser without executing embedded markup.
C:The action is useful for the separate capability to teach a classifier how to label borderline requests consistently because diverse demonstrations make the intended decision boundary clearer than an isolated easy example, provided labels are accurate; however, the required test outcomes are to stop verbose tool payloads from consuming most of a long-running agent's context and to safely render generated text in a browser without executing embedded markup.
E:The action is useful for the separate capability to improve a prompt while preserving performance on previously solved cases because controlled prompt iteration distinguishes genuine quality gains from regressions or overfitting to one attractive example; however, the required test outcomes are to stop verbose tool payloads from consuming most of a long-running agent's context and to safely render generated text in a browser without executing embedded markup.
Question 56
A change request for a production incident affecting a Claude endpoint combines two needs: locate which component of a multi-tool answer introduced the wrong value, and distinguish a syntactically valid but low-quality answer from a transport outage. Which implementation pair addresses both, not just one?
- Perform both steps: (1) Inspect the HTTP 429 status and rate-limit metadata before retrying; (2) compare versioned prompts and representative eval cases against model responses.
- Perform both steps: (1) Inspect the HTTP 429 status and rate-limit metadata before retrying; (2) trace the api credential and access permissions for the 401 or 403 response.
- Perform both steps: (1) Examine correlated request, tool, and transformation traces stage by stage; (2) trace the api credential and access permissions for the 401 or 403 response.
- Perform both steps: (1) Examine correlated request, tool, and transformation traces stage by stage; (2) compare versioned prompts and representative eval cases against model responses.
Explanation
Correct Answer
D:The first change enables the team to locate which component of a multi-tool answer introduced the wrong value because structured traces reveal whether retrieval, tool execution, postprocessing, or generation caused the defect; reading only the final text cannot isolate it; the second addresses the separate need to distinguish a syntactically valid but low-quality answer from a transport outage because a successful HTTP call only proves the protocol worked; quality regressions require output-level evidence and controlled comparisons.
Incorrect Answers
A:The second step meets the requirement to distinguish a syntactically valid but low-quality answer from a transport outage because a successful HTTP call only proves the protocol worked; quality regressions require output-level evidence and controlled comparisons; the first step would only help to distinguish an upstream throttling response from a malformed model answer because a rate limit is a request-layer capacity problem; the response needs controlled pacing rather than a prompt rewrite, leaving the separate requirement to locate which component of a multi-tool answer introduced the wrong value unmet.
B:Both steps target different results—distinguish an upstream throttling response from a malformed model answer and identify why a request is rejected before any inference happens—because a rate limit is a request-layer capacity problem; the response needs controlled pacing rather than a prompt rewrite and unauthorized access fails at the transport or authorization boundary, not because Claude misunderstood the prompt; the required capabilities are instead to locate which component of a multi-tool answer introduced the wrong value and to distinguish a syntactically valid but low-quality answer from a transport outage.
C:The first step is appropriate for the requirement to locate which component of a multi-tool answer introduced the wrong value because structured traces reveal whether retrieval, tool execution, postprocessing, or generation caused the defect; reading only the final text cannot isolate it; the other step instead aims to identify why a request is rejected before any inference happens because unauthorized access fails at the transport or authorization boundary, not because Claude misunderstood the prompt, which does not meet the second requirement to distinguish a syntactically valid but low-quality answer from a transport outage.
Question 57
An architecture review of a tool-calling Claude workflow asks how to make reference material available for clients to read as context. Which action is the best fit?
- Expose the material as an MCP resource with a stable identifier.
- Package the connector behind an MCP server.
- Define a descriptive tool name and validated input schema.
- Validate model-supplied tool arguments against schema and domain rules.
Explanation
Correct Answer
A:To make reference material available for clients to read as context in this tool-calling Claude workflow, resources represent consumable data, contrasting with tools that invoke an operation or prompts that package reusable instructions; the proposed implementation directly supplies that capability.
Incorrect Answers
B:The proposed alternative helps to share one integration with several compatible AI clients or agent runtimes because MCP standardizes capability exposure across clients, reducing duplicate integration logic compared with proprietary tool schemas everywhere; that mechanism does not meet the different requirement to make reference material available for clients to read as context.
C:The proposed alternative helps to expose a business lookup with predictable typed arguments because well-described schemas help Claude select a capability and emit usable arguments; the backend still validates authorization and meaning; that mechanism does not meet the different requirement to make reference material available for clients to read as context.
D:The proposed alternative helps to reject a date or identifier that violates business constraints before a service call because even syntactically correct generated inputs can violate account scope or allowed values, so execution must validate both levels; that mechanism does not meet the different requirement to make reference material available for clients to read as context.
Question 58
Two approval criteria are still open for an internal coding-assistant workflow: the team must recover a multi-hour agent workflow after an infrastructure interruption and coordinate explicit graph transitions among multiple model and tool nodes. Which two implementation actions satisfy them? Choose TWO.
- Use isolated branch state and a controlled reconciliation step.
- Choose a graph-based orchestration framework when state transitions justify it.
- Alternate model tool-use requests with tool results until completion.
- Give an agent a bounded objective and a tool-selection loop.
- Persist checkpoints for tool outputs and accepted decisions.
Explanation
Correct Answers
B:To coordinate explicit graph transitions among multiple model and tool nodes as part of a design that must also recover a multi-hour agent workflow after an infrastructure interruption, graph orchestration clarifies conditional edges and resumability; it adds complexity when a short linear function would suffice; this is why the action is one of the two required selections.
E:To recover a multi-hour agent workflow after an infrastructure interruption as part of a design that must also coordinate explicit graph transitions among multiple model and tool nodes, durable checkpoints let the orchestrator resume from a known stage rather than rerunning costly or side-effecting actions; this is why the action is one of the two required selections.
Incorrect Answers
A:This alternative serves the separate goal to stop two parallel branches from overwriting the same mutable record because explicit merge rules are required when outputs converge; concurrent mutation without arbitration causes nondeterministic results; the required outcomes are instead to recover a multi-hour agent workflow after an infrastructure interruption and to coordinate explicit graph transitions among multiple model and tool nodes.
C:This alternative serves the separate goal to plan another step only after inspecting the result of a requested external operation because the loop returns observations to the model so subsequent decisions are informed by actual tool output rather than imagined execution; the required outcomes are instead to recover a multi-hour agent workflow after an infrastructure interruption and to coordinate explicit graph transitions among multiple model and tool nodes.
D:This alternative serves the separate goal to pursue an open-ended investigation where the next action depends on intermediate evidence because agentic control lets the model choose among permitted steps as observations arrive; budgets and termination conditions still bound its autonomy; the required outcomes are instead to recover a multi-hour agent workflow after an infrastructure interruption and to coordinate explicit graph transitions among multiple model and tool nodes.
Question 59
A production readiness review of a hosted document-processing service finds that the team cannot yet keep application policy separate from a user's question or filter risky user-visible content according to application policy. Which paired changes directly remedy the two gaps?
- Perform both steps: (1) Store and retrieve message history by authenticated tenant and conversation ID; (2) apply content checks at suitable input and output boundaries.
- Perform both steps: (1) Place stable application rules in system instructions and untrusted requests in user content; (2) route claude requests through a trusted backend service.
- Perform both steps: (1) Place stable application rules in system instructions and untrusted requests in user content; (2) apply content checks at suitable input and output boundaries.
- Perform both steps: (1) Store and retrieve message history by authenticated tenant and conversation ID; (2) route claude requests through a trusted backend service.
Explanation
Correct Answer
C:The first change enables the team to keep application policy separate from a user's question because instruction placement preserves the distinction between developer intent and user-supplied data; concatenating everything blurs that boundary; the second addresses the separate need to filter risky user-visible content according to application policy because moderation is a defense layer applied around generation, not a substitute for authenticated tool access or scoped retrieval.
Incorrect Answers
A:The second step meets the requirement to filter risky user-visible content according to application policy because moderation is a defense layer applied around generation, not a substitute for authenticated tool access or scoped retrieval; the first step would only help to prevent chat transcripts from different accounts entering the same Claude request because session hygiene requires server-enforced isolation; model instructions cannot repair a history assembled from the wrong user, leaving the separate requirement to keep application policy separate from a user's question unmet.
B:The first step is appropriate for the requirement to keep application policy separate from a user's question because instruction placement preserves the distinction between developer intent and user-supplied data; concatenating everything blurs that boundary; the other step instead aims to keep a browser application from exposing the secret used for model calls because a backend controls credentials and access policy while the UI sends user input through an authenticated application endpoint, which does not meet the second requirement to filter risky user-visible content according to application policy.
D:Both steps target different results—prevent chat transcripts from different accounts entering the same Claude request and keep a browser application from exposing the secret used for model calls—because session hygiene requires server-enforced isolation; model instructions cannot repair a history assembled from the wrong user and a backend controls credentials and access policy while the UI sends user input through an authenticated application endpoint; the required capabilities are instead to keep application policy separate from a user's question and to filter risky user-visible content according to application policy.
Question 60
The technical owner of an agent access-control assessment has to stop frontend users from receiving unrestricted backend tool credentials. A separate requirement is to keep Claude API credentials out of repository source and deployment images. Which two-part remediation covers both?
- Perform both steps: (1) Validate or moderate final content before release; (2) retrieve keys through a managed secrets facility at runtime.
- Perform both steps: (1) Use a server-side workload identity or scoped service credential; (2) retrieve keys through a managed secrets facility at runtime.
- Perform both steps: (1) Use a server-side workload identity or scoped service credential; (2) record authenticated approval events with actor, scope, and request correlation.
- Perform both steps: (1) Validate or moderate final content before release; (2) record authenticated approval events with actor, scope, and request correlation.
Explanation
Correct Answer
B:The first change enables the team to stop frontend users from receiving unrestricted backend tool credentials because service identities tie privileges to a runtime workload and avoid exposing long-lived secrets to browsers or untrusted users; the second addresses the separate need to keep Claude API credentials out of repository source and deployment images because central secret handling improves rotation and auditing and avoids accidental credential disclosure through source control.
Incorrect Answers
A:The second step meets the requirement to keep Claude API credentials out of repository source and deployment images because central secret handling improves rotation and auditing and avoids accidental credential disclosure through source control; the first step would only help to prevent generated advice that violates the product's publication rules from being shown because output checks catch issues emerging during generation; they complement rather than replace input controls and tool restrictions, leaving the separate requirement to stop frontend users from receiving unrestricted backend tool credentials unmet.
C:The first step is appropriate for the requirement to stop frontend users from receiving unrestricted backend tool credentials because service identities tie privileges to a runtime workload and avoid exposing long-lived secrets to browsers or untrusted users; the other step instead aims to demonstrate who approved a privileged tool action and when because an audit trail binds authorization to a specific action, making privilege use reviewable without treating model-generated text as proof of consent, which does not meet the second requirement to keep Claude API credentials out of repository source and deployment images.
D:Both steps target different results—prevent generated advice that violates the product's publication rules from being shown and demonstrate who approved a privileged tool action and when—because output checks catch issues emerging during generation; they complement rather than replace input controls and tool restrictions and an audit trail binds authorization to a specific action, making privilege use reviewable without treating model-generated text as proof of consent; the required capabilities are instead to stop frontend users from receiving unrestricted backend tool credentials and to keep Claude API credentials out of repository source and deployment images.